Hugo Kubinyi, www.kubinyi.

de

QSAR Parameters

Hugo Kubinyi
Germany

E-Mail kubinyi@t-online.de
HomePage www.kubinyi.de

Hugo Kubinyi, www.kubinyi.de

What is QSAR ?
QSAR (quantitative structure-activity relationships)
includes all statistical methods, by which biological
activities (most often expressed by logarithms of
equipotent molar activities) are related with structural
elements (Free Wilson analysis), physicochemical
properties (Hansch analysis), or fields (3D QSAR).
Classical QSAR analyses (Hansch- and Free Wilson
analyses) consider only 2D structures. Their main field
of application is in substituent variation of a common
scaffold.
3D-QSAR analysis (CoMFA) has a much broader scope.
It starts from 3D structures and correlates biological
activities with 3D-property fields.
Hugo Kubinyi, www.kubinyi.de

Drug Action: From Experience to Theory to Rules

1900, H. H. Meyer and C. E. Overton: lipoid theory of
narcosis
1930‘s, L. Hammett: electronic sigma constants
1964, C. Hansch and T. Fujita: QSAR
1984, P. Andrews: affinity contributions of functional
groups
1985, P. Goodford: GRID (hot spots at protein surface)
1988, R. Cramer: 3D QSAR
1992, H.-J. Böhm: LUDI interaction sites, docking,
scoring
1997, C. Lipinski: bioavailability rule of five
1998, Ajay, W. P. Walters and M. A. Murcko; J. Sadowski
and H. Kubinyi: drug-like character

Hugo Kubinyi, www.kubinyi.de

JACS 86, 1616 (1964)

 Hugo Kubinyi, www.kubinyi.de

Drug Transport and Drug Receptor Interaction

The “random walk” process

drug receptor
aqueous phases and (binding site)
lipophilic barriers

biological activity = f (transport + binding) =

- k1 (lipo)² + k2 (lipo) + k3 (pol) + k4 (elec) + k5 (ster) + k6

Hugo Kubinyi, www.kubinyi.de

Specifics of Drug Action

Lipophilicity and dissociation / ionization are responsible
for transport and distribution of drugs in biological
systems.
The geometric fit and the complementarity of the surface
3D properties of a ligand are responsible for its affinity to
a binding site.

- Which conformation is the biologically active conforma-

tion ?
conformation in vacuo
conformation in the crystal
conformation in aqueous solution
conformation at the binding site
 Hugo Kubinyi, www.kubinyi.de

Basic Requirements in QSAR Studies

- all analogs belong to a congeneric series

- all analogs exert the same mechanism of action

- all analogs bind in a comparable manner

- the effects of isosteric replacement can be predicted

- binding affinity is correlated to interaction energies

- biological activities are correlated to binding affinity

Hugo Kubinyi, www.kubinyi.de

Molecular Properties and Their Parameters

Molecular Property Corresponding Interaction Parameters
Lipophilicity hydrophobic interactions log P, π , f, RM , χ

Polarizability van-der-Waals interactions MR, parachor, MV

Electron density ionic bonds, dipol-dipol σ , R, F, κ , quantum

interactions, hydrogen chemical indices
bonds, charge transfer
interactions

Topology steric hindrance Es, r v, L, B,

geometric fit distances, volumes
X
Hammett equation ρσ = log kRX - log kRH
Hugo Kubinyi, www.kubinyi.de

QSAR Models - Hansch model

(property-property relationship)
Definition of the lipophilicity
parameter π
πX = log PRX - log PRH

Linear Hansch model

Log 1/C = a log P + b σ + c MR + ... + k

Nonlinear Hansch models

log 1/C = a (log P)2 + b log P + c σ + ... + k
log 1/C = a π2 + b π + c σ + ... + k
log 1/C = a log P - b log (ßP + 1) + c σ + ... + k

Hugo Kubinyi, www.kubinyi.de

Free-Wilson model (structure-property

relationship)
X1 X2
X3
log 1/C = Σ ai + µ

ai = substituent group contributions

µ = activity contribution of reference compound

Mixed Hansch/Free-Wilson model

log 1/C = a (log P)2 + b log P + c σ + ... + Σ ai + k

log 1/C = a log P - b log (ßP + 1) + c σ + ... + Σ ai + k
 Hugo Kubinyi, www.kubinyi.de

Lipophilicity

Hydrophobic interaction between a drug and a binding

site at a receptor
Definition of Partition Coefficients
P = corg/caq (n-octanol/water system)

Hugo Kubinyi, www.kubinyi.de

Octanol/Water Partition Coefficients of Carbamates

J. B. Houston et al., J. Pharmacol. Exp. Ther. 189, 244 (1974)

R-OCONH2 P log P ∆ log P = πCH2

Methyl 0.22 -0.66
} 0.51
Ethyl 0.70 -0.15
} 0.51
Propyl 2.3 0.36
} 0.49
Butyl 7.1 0.85
} 0.50
Pentyl 22.5 1.35
} 0.50
Hexyl 70.8 1.85
} 0.51
Heptyl 230 2.36
} 0.49
Octyl 700 2.85
sec-Butyl 4.5 0.65 -0.20 *)
tert-Butyl 3.0 0.48 -0.37 *)
*) relative to n-butyl carbamate
Hugo Kubinyi, www.kubinyi.de

n-Octanol/Water as a Standard System

- membrane analogous structure
- hydrogen bond donor and acceptor
- practically insoluble in water
- no desolvation on transfer into organic phase
- very low vapor pressure
- transparent in the UV region
- large data base of log P values
Additivity Principle of π Values (C. Hansch, 1964)
πX = log PR-X - log PR-H
The lipophilicity parameter π is an additive, constitutive
molecular parameter; compare the Hammett Equation:
ρσX = log KR-X - log KR-H

Hugo Kubinyi, www.kubinyi.de

π Values of Aromatic Substituents

Substituent πmeta πpara πBenzene
H 0.00 0.00 0.00
CH3 0.51 0.52 0.56
Cl 0.76 0.70 0.71
Br 0.94 1.02 0.86
OH -0.49 -0.61 -0.67
OCH3 0.12 -0.04 -0.02
NO2 0.11 0.24 -0.28

! cf.: Lipophilicity contribution of an aliphatic

CH2 group = 0.50
 Hugo Kubinyi, www.kubinyi.de

Toxicity of Substituted Benzoic Acids in Mosquito Larvae

all compounds, σ
log 1/C = 1.069 σ + 1.780
(n = 14; r = 0.711; s = 0.427)
without the 4-nitro analog, σ (Hansen, 1962)
log 1/C = 1.454 σ + 1.787
(n = 13; r = 0.918; s = 0.243)

all compounds, π (Hansch and Fujita, 1964)

log 1/C = 0.535 π + 1.602
(n = 14; r = 0.969; s = 0.151)
intercorrelation π vs. σ (without 4-nitro analog)
r = 0.91

Hugo Kubinyi, www.kubinyi.de

Toxicity of Substituted Benzoic Acids in Mosquito Larvae

Hugo Kubinyi, www.kubinyi.de

Calculation of the Log P Value of m-Chloro-

toluene
Cl CH3

log P = log PBenzene + πCl + πMe

= 2.13 + 0.71 + 0.56 = 3.40
log P = log PToluene + πmeta-Cl
= 2.69 + 0.76 = 3.45
log P = log PChlorobenzene + πmeta-Me
= 2.84 + 0.51 = 3.35
log Pexp = 3.28

Hugo Kubinyi, www.kubinyi.de

Nonadditivity of π Values
- intramolecular hydrogen bonds
- ortho effects (phenols)
- polysubstituted aromatic compounds
- conjugation (push pull effect)
- heterocycles
- cyclophanes
CH3

CH3
surface area 150.3 196.6
exptl. Log P 3.15 2.33
Hugo Kubinyi, www.kubinyi.de

Wrong Application („Failure“) of the π Concept

πX = log PR-X - log PR-H (correct)

log P = Σπi (wrong)

log PC6H5CH3 = log PC6H6 + πCH3 (correct)

log PC6H5CH2CH2C6H5 = 2 log PC6H6 + 2 πCH3

= 4.26 + 1.00 = 5.26 (wrong)
1,2-diphenylethane, exp. log P = 4.79

Hugo Kubinyi, www.kubinyi.de

Hydrophobic Fragmental Constants f

R. Rekker, The Hydrophobic Fragmental Constant, Elsevier,
Amsterdam 1977; R. Rekker. Eur. J. Med. Chem. 14, 479 (1979)
log P = Σ aifi (R. Rekker, 1973)

Fragment Rekker constant f Leo constant f π benzene

H 0.175 0.23 0.00
CH3 0.702 0.89 0.56
>CH2 0.530 0.66
>CH- 0.235 0.43
C6 H5 1.886 1.90 1.96
C6 H4 1.688
C6 H3 1.431
log P C6 H5CH2 CH2 C6 H5 = 2 fC6 H5 + 2 fCH2
= 3.77 + 1.06 = 4.83 (correct)
Hugo Kubinyi, www.kubinyi.de

Fragment Rekker constant f Leo constant f π benzene

Aliph. Arom. Aliph. Arom.
F -0.462 0.399 -0.38 0.37 0.14
Cl 0.061 0.922 0.06 0.94 0.71
Br 0.270 1.131 0.20 1.09 0.86
I 0.587 1.448 0.60 1.35 1.12
OH -1.491 -0.343 -1.64 -0.40 -0.67
-O- -1.581 -0.433 -1.81 -0.57
COOH -0.954 -0.093 -1.09 -0.03 -0.32
NH2 -1.428 -0.854 -1.54 -1.00 -1.23
NH -1.825 -0.964 -2.11 -1.03
NO2 -0.939 -0.078 -1.26 -0.02 -0.28
CONH2 -1.970 -1.109 -2.18 -1.26 -1.49
>C=O -1.703 -0.842 -1.90 -0.32
CF3 0.757 1.331 0.88
C≡≡N -1.066 -0.205 -1.28 -0.34 -0.57
SH 0.000 0.620 0.39
-S- -0.510 0.110 -0.79 0.03

Hugo Kubinyi, www.kubinyi.de

Experimental Determination of Log P Values

- Shake flask method
- Reversed phase thin layer chromatography
- High performance liquid chromatography (HPLC)

High Performance Liquid Chromatography (HPLC)

log P = a log k´ + const. k´ = (tr - to)/to
different standard compounds, n-octanol-saturated,
persilylated CORASIL C-18®, pH 7 phosphate buffer

log P = 1,025 (±0.06) log k´ + 0.797

(n = 33; r = 0.987; s = 0.127)

S. Unger et al., J. Pharm. Sci. 67, 1364 (1978)

Hugo Kubinyi, www.kubinyi.de

Lipophilicity Estimation Software

CLogP (Daylight Chemical Information Systems)
http://www.daylight.com/daycgi/clogp
KowWin (Syracuse Research Corp., SRC; W. M. Meylan
and P. H. Howard, J. Pharm. Sci. 84, 83-92 (1995))
http://esc.syrres.com/interkow/kowdemo.htm
Interactive Analysis‘s LogP Program (Mark Parham;
includes Lipinski numbers)
http://www.logp.com/main.html
Neuro-Heuristic Program ALogPS (Igor Tetko, Lausanne)
http://www.vcclab.org
Experimental LogP values (SRC, n = 13250)
http://esc.syrres.com/interkow/PhysProp.htm

Hugo Kubinyi, www.kubinyi.de

Polarizability Parameters
Molar volume, Molar Refractivity, Parachor

MW n2 - 1
MV = MR = . MW
d n2 + 2 d

MW
PA = γ 1/4 .
d
d = density; n = refraction index;
γ = surface tension

(MR is most often scaled by a factor of 0.1)

Hugo Kubinyi, www.kubinyi.de

Binding of Neutral Compounds to BSA

Log 1/C = 0.751 (±0.07) log P + 2.300
(n = 42; r = 0.960; s = 0.159)

Log 1/C = 0.024 (±0.02) MR + 2.901

(n = 42; r = 0.307; s = 0.536)
Binding of Phenyl ß-D-Glucosides to
Concanavalin A
Log M50 = 0.971 (±0.56) π + 2.37
(n = 19; r = 0.664; s = 0.095)

Log M50 = 0.019 (±0.003) MR + 2.23

(n = 19; r = 0.954; s = 0.038)

Hugo Kubinyi, www.kubinyi.de

4-Hydroxyquinoline Carboxylic Acids,

Inhibition of Malate Dehydrogenase OH
COOH
pI50 = 0.70 (±0.17) MR + 2.29
(n = 13; r = 0.939; s = 0.315) R N
4-Hydroxyquinoline Carboxylic Acids,
Inhibition of Ascites Cell Respiration
pI50 = 0.46 (±0.11) π + 3.22
(n = 14; r = 0.933; s = 0.280)

Binding of N-Acyl L-Amino Acid Methyl Esters

R1CONH-CH(R2)-COOMe to Chymotrypsin
Log 1/Km = 0.082 (±0.02) MR1 + 1.382 (±0.87) π2 – 3.876
(n = 21; r = 0.934; s = 0.331)
 Hugo Kubinyi, www.kubinyi.de

Electronic Parameters
Hammett Equation ρσ = log KRX - log KRH
H2O
COOH COO
- +H
+
ρ = 1.00
R 25° C R
H2O
CH2COOH CH2COO
- +H
+
ρ = 0.49
R 25° C R
H2 O
- +
OH O +H ρ = 2.11
R 25° C R
H2 O
+ +
NH2 NH2 + H ρ = 2.77
R 25° C R

Hugo Kubinyi, www.kubinyi.de

Hammett Equation ρσ = log kRX - log kRH

COOEt + OH - alkaline hydrolysis

ρ = 2.43
R in 85% EtOH, 30° C

acid-catalyzed
COOH + CH3OH
ester formation, 25° C ρ = - 0.23
R

- alkylation of phenols
O + C2H5 I in EtOH, 42.5° C ρ = - 0.99
R

CH2Cl + OH - nucleophilic replacement

ρ = - 0.33
R in water, 30° C

benzoylation of anilines
NH2 + C6H5COCl
in benzene, 25° C ρ = - 2.78
R
Hugo Kubinyi, www.kubinyi.de

Calculation of pKa Values pKa R-X = pKa R-H - ρσ

pKa value of 3,5-dinitro-4-methyl-benzoic acid

(pKa benzoic acid = 4.20)
experimental value = 2.97
calculated value = 4.20 - (0.71 - 0.17 + 0.71) = 2.95
pKa value of m-hydroxybenzoic acid
(pKa benzoic acid = 4.20, phenol = 9.92;
ρphenols = 2.23; -OH: σmeta = 0.12; -COOH: σmeta = 0.37;
-COO-: σmeta = -0.10)
experimental values = 4.06 and 9.92
calculated for the carboxylate group
= 4.20 - 0.12 = 4.08
calculated for the phenolic OH group
= 9.92 - 2.23·(-0.10) = 10.14

Hugo Kubinyi, www.kubinyi.de

pKa value of 3-methoxy-5-nitro-aniline

(pKa aniline = 4.57; ρaniline = 2.81)
experimental value = 2.13
calculated value = 4.57 - 2.81 (0.12 + 0.71) = 2.24
pKa values of chloro-, trichloro- and cyanoacetic acid
(pKa RCOOH = 4.66; ρRCOOH = 1.62; σ* for -CH2Cl,
-CCl3 and -CH2C≡ ≡N = 1.05, 2.65 and 1.30; σ* =
aliphatic σ value)
chloroacetic acid, experimental value = 2.85
calculated value = 4.66 - 1.62·(1.05) = 2.96
trichloroacetic acid, experimental value = 0.70
calculated value = 4.66 - 1.62·(2.65) = 0.37
cyanoacetic acid, experimental value = 2.45
calculated value = 4.66 - 1.62·(1.30) = 2.55
Hugo Kubinyi, www.kubinyi.de

Quantum Mechanical Descriptors

Atom partial charges:
Mulliken population analysis (orbital population)
ESP charges (mapping EP to atom locations)
Dipole moment:
strength and orientation behavior of a molecule in
an electrostatic field
E
HOMO / LUMO (“frontier orbital theory“):
HOMO = energy of highest occupied LUMO
molecular orbital, “nucleophilicity“ HOMO
LUMO = energy of lowest unoccupied
molecular orbital, “electrophilicity“
Superdelocalizability:
estimate for the reactivity of positions in donor acceptor
aromatic hydrocarbons

Hugo Kubinyi, www.kubinyi.de

Steric Parameters
Es values (acid-catalyzed hydrolysis of RCOOR´,
based on CH3COOR´)
Es = log (k/ko)A
Van-der-Waals radii of symmetric substituents
(rv (av) = mean radius)
Es = - 1.839 rv (av) + 3.484
(n = 6; r = 0.996; s = 0.132)
Charton‘s steric substituent parameter
(rv = minimal van-der-Waals radius)
υX = rvX - rvH = rvX - 1.20
Es = - 2.062 (±0.86) υ - 0.194 (±0.10)
(n = 104; r = 0.978; s = 0.250)
Hugo Kubinyi, www.kubinyi.de

Hancock Esc values (corrected for the number of

α-hydrogen atoms, nH)

Esc = Es + 0.306 (nH - 3)

Fujita Esc values for unsymmetrical substituents

CR1R2R3: weighted sum of the Esc values of

R1, R2 and R3

Verloop‘s STERIMOL parameters

L, B1, B2, B3, and B4

revised parameters L, B1, and B5

Hugo Kubinyi, www.kubinyi.de

Sterimol
Parameters
(Arië Verloop)

A graphical definition
of second generation
STERIMOL parameters
(Verloop, 1987), using
the substituent
CH2CH2COCH3 as an
example; L is the length,
B1 the minimal width
and B5 a newly defined
maximal width
 Hugo Kubinyi, www.kubinyi.de

Kier-Hall Molecular Connectivity Indices iχ

0 1 2 3
Bond-path indices
for isopentane Path
= electron-weighted
subgraph counts

0χ = Σ (# σ-electrons of i)-0.5

1χ = Σ (0χ(i).0χ(j))-0.5
(over all bonds ij)
Cluster
... etc.
0χ 1χ 2χ 3χ 3χ
P, C

Hugo Kubinyi, www.kubinyi.de

Parameter Tables and Log P Data Bases

C. Hansch et al., J. Med. Chem. 16, 1207 (1973); J. Med.
Chem. 20, 304 (1977); 284 aromatic substituents, π, σmeta,
σpara, F, R and MR
C. Hansch and A. Leo, Substituent Constants for Correlation
in Chemistry and Biology, Wiley, New York, 1979. All known
substituent parameters; about 15,000 log P values
C. Hansch, A. Leo, and D. Hoekman, Exploring
QSAR. Hydrophobic, Electronic, and Steric Constants,
American Chemical Society, Washington, 1995.
R. Todeschini and V. Consonni, Handbook of Molecular
Descriptors, Methods and Principles in Medicinal Chemistry,
Volume 11, R. Mannhold, H. Kubinyi and H. Timmerman,
Eds., Wiley-VCH, Weinheim, 2000.

MedChem and BioByte Databases, Program DRAGON

 Hugo Kubinyi, www.kubinyi.de

Program DRAGON

Roberto Todeschini

free download of the

DRAGON program from
www.disat.unimib.it/
chm/Dragon.htm

calculates 1,481 molecular

descriptors, distributed
in 18 blocks, from MOL,
SDF or Hyperchem files.

Hugo Kubinyi, www.kubinyi.de

The Software DRAGON

R. Todeschini, © Talete srl
calculates 1,481 molecular descriptors, distributed in 18 blocks.
General features of the version 2.1:
Each molecule has to be represented by 3D geometrical coordinates,
hydrogen included.
The maximum number of atoms (hydrogen atoms included) is 150.
Calculations are quite fast; however, several condensed or adjacent
rings can slow down the calculations.
The total number of molecules was extended to 1,000.
Features of the commercial version:
up to 2,500 molecules (instead of 1,500),
maximum number of 300 atoms, including hydrogen (instead of 150)
maximum number of 100 y responses (instead of 20)
number of molecular descriptors: 1,486 (instead of 1,481;
new models for the calculation of experimental properties)
loading of SMILES format (not possible in the Web version)
Hugo Kubinyi, www.kubinyi.de

Hugo Kubinyi, www.kubinyi.de